Development of Techniques for IN Vivo Investigation of the Function of Identified Neurons We seek to understand how basal ganglia circuitry contributes to action control and learning. To this end, we want to gain an understanding of the activity of specific types of neurons within this circuitry during the learning, planning and performance of actions and action sequences. However, it has been difficult to determine the function of subclasses of closely-related neurons within basal ganglia subregions, as electrophysiological measures are not sufficient to allow neuronal subclass identification. For example, the medium spiny projection neurons (MSNs) in striatum send inhibitory output via two pathways, the direct and indirect pathways. The action potential properties of these two MSN subtypes are very similar, and thus investigators have had difficulty determining how changes in the activity of the different subtypes relate to actions. To address this problem, we worked with Drs. Rui Costa of the Champalimaud Neuroscience Institute and Steven Vogel of the Laboratory of Molecular Physiology at NIAAA to develop a technique for in vivo fluorimetry using fiber optics to perform Time Correlated Single Photon Counting (TCSPC) fluorimetry in the striatum in vivo. Coupling this technique with selective expression of a fluorescent calcium indicator protein (GCaMP3) in either direct or indirect pathway MSNs allowed us to measure calcium transients reflective of neuronal activation. We found co-activation of neurons in the 2 pathways during initiation of movement sequences. We are now working to improve the sensitivity of detection of calcium transients, and extending this approach to other neuronal subtypes. In recent experiments we have employed the GCaMP5 indicator protein that has enhanced dynamic range in comparison to GCaMP3. The calcium transients detected with this fluorescent protein were indeed larger in amplitude than those observed using GCaMP3 when expressed in a similar population of MSNs. Thus, we are confident that this increased signal/noise ratio will allow us to detect calcium signals in neurons with greater sensitivity, perhaps even allowing us to measure activity in neuronal populations sparser than the MSNs. We have also conducted preliminary studies using the fluorimetry/TCSPC system to examine GCaMP-reporting of calcium transients in midbrain dopaminergic neurons. Our findings to date indicate that responses to stressful environmental stimuli can be detected in populations of midbrain neurons using this approach. We are currently determining how well the fiber optic fluorimetry system will work for detection of fluorescence generated by other proteins and molecular reporters expressed in striatal neurons. We can detect stable fluorescent signals generated by green fluorescent protein, and we have also been able to detect fluorescence generated by the membrane fusion detector synaptopHluorin. We hope to adapt this system for real time detection of changes in a variety of biomolecules in deep brain structures during ongoing behavior. Early Life Ethanol Exposure Disrupts Habitual Control of Operant Behavior Fetal alcohol exposure in humans produces Alcohol-Related Neurodevelopmental Disorder (ARND) that includes an array of neurological and behavioral abnormalities. Among these are hyperactivity and attentional problems, as well as cognitive impairments that could stem from fetal alcohol effects on corticostriatal circuitry. However, little is known about the mechanisms underlying these aspects of ARND, or the alterations in corticostriatal circuitry brought about by fetal alcohol exposure. Thus, we have examined striatally-controlled operant behavior and striatal physiology in a mouse model of fetal alcohol exposure. Mice were exposed to ethanol throughout gestation and during the early postnatal period, mimicking a full 3 trimester exposure in humans. The level of ethanol exposure produced relatively moderate blood alcohol concentrations, near the legal limit for intoxication. We term this model Gestational Ethanol Exposure (GEE). GEE and control mice were trained and tested in a food-rewarded instrumental lever pressing task as adults. The GEE mice did not use a habitual strategy for task performance, as evidenced by strong decreases in lever pressing after sensory-specific satiety devaluation of the food reward under conditions in which control mice show no such decrease. This impaired habit formation persisted even after several weeks of training using a random interval training schedule that normally fosters robust insensitivity to reward devaluation. We sought to determine what physiological changes might underlie impaired habit learning in the GEE mice. Using in vivo multi-electrode electrophysiological recording, we observed that the average action potential firing rate of neurons in the dorsolateral striatum (a region that controls habits) was increased in the GEE animals relative to control mice. We also observed that the number of neurons showing altered firing associated with lever presses decreased over the course of training and devaluation testing in GEE, but not control, mouse striatum. Finally, we found that abnormally high numbers of striatal neurons in GEE mice were activated in association with lever pressing when animals were undergoing training using the random interval schedule, and this pattern shifted to abnormally low activity during devaluation testing. Dorsolateral striatal neurons appear to encode information in a value-sensitive way in GEE mice under conditions in which neurons in control animals show no such sensitivity. These circuitry changes likely maintain behavioral sensitivity to outcome even under conditions when insensitivity is a more adaptive strategy. If this sort of change occurs in humans, it might lead to rapid shifts in attention to new goals, with an inability to prioritize stimulus-action-outcome relationships. In brain slice electrophysiological recordings we found a decrease in GABAergic synaptic transmission involving reduced GABA release onto MSNs in dorsolateral striatum. Experiments examining presynaptic modulation involving the CB1 cannabinoid receptor indicated abnormal tonic activation of the receptor by the endocannabinoid 2-arachidonoyl glycerol in GEE, relative to control mouse MSNs. In slices from control mice, application of inhibitors of endocannabinoid-degrading enzymes decreased striatal GABAergic transmission to an extent similar to that observed in GEE mouse slices. We reasoned that if tonic endocannabinoid signaling contributed to the loss of habitual behavior in the GEE mice, then we should be able to phenocopy the behavioral change with tonic inhibition of endocannabinoid degradation. Indeed, we found that treatment with inhibitors of endocannabinoid degradation in adult control mice led to a loss of habitual instrumental lever pressing that closely resembled that seen in adult GEE mice. This finding indicates that abnormally high endocannabinoid tone is a viable mechanism for the impaired habit formation in GEE mice, and that this effect only requires tonic endocannabinoid actions in the adult animal, providing intriguing possibilities for reversing the GEE effects by reducing these tonic actions. Our work indicates that moderate alcohol exposure early in life leads to impaired habitual control well into adulthood. This cognitive/behavioral impairment has similarities to some of the behavioral changes observed in humans exposed to fetal alcohol. It will be interesting to determine if habit formation is altered in individuals diagnosed with ARND. In addition, the changes in endocannabinoid function that we have characterized could be used to develop markers, and perhaps treatments, for ARND.